Training · Cohort

The Constrained Energy Model: More Exercise Won’t Burn More

Physical activity explains less than 9% of the variation in how many calories you burn per day. The other 91%? Your body decides — and it’s not taking requests.

Listen while you read · FitChef Audio
“The most active people in the study didn’t burn detectably more calories per day than moderately active people. Their bodies quietly redirected the energy — spending less on other biological processes to keep total expenditure roughly constant.”
— Pontzer et al. 2016 · 332 adults across 5 populations

Physical activity — all of it, from your morning walk to your evening spin class — explains somewhere between seven and nine percent of the variation in how many calories people burn per day. Not seventy. Not fifty. Single digits.

That number comes from a study of 332 adults across five countries, measured with doubly labeled water — the most accurate method science has for tracking real energy expenditure in free-living humans. No self-reported food diaries. No gym machine estimates. Actual metabolic measurement, over seven days.

And what Herman Pontzer and his research team found in that data, published in Current Biology in 2016, rewrites the arithmetic most people use when they step off a treadmill and onto a scale.

Researchers tracked 332 adults across five countries using the gold standard for measuring calorie burn — and found that above a moderate activity threshold, the relationship between exercise and total daily energy expenditure was statistically indistinguishable from zero. More movement didn’t mean more calories burned. The budget had already been spent.
Pontzer et al. 2016 · Current Biology · 332 adults, 5 populations
Key takeaways

Your body runs on a calorie budget that doesn't expand — no matter how many spin classes you add. The calories you burn through exercise get quietly reshuffled, not stacked on top.

  • The study tracked 332 adults across five countries using the gold-standard method for measuring calorie burn — and found that physical activity explains less than 9% of the variation in how many calories people burn per day.
  • Your body size — specifically how much muscle you carry — is the single biggest factor in your daily calorie burn, accounting for about 52% of the variation. Your workout accounts for single digits.
  • People with more body fat face a steeper metabolic penalty: their bodies compensate for exercise calories more aggressively than leaner people doing the exact same workout.
  • After controlling for total activity, every extra minute of vigorous exercise was associated with 17 to 18 fewer total calories burned per day — not more.
  • The researchers found that exercise may benefit health not by burning extra calories, but by redirecting the body's existing calorie budget away from inflammation and toward more beneficial processes.

The Calorie Ceiling Nobody Told You About

The researchers plotted total daily energy expenditure against physical activity levels for all 332 participants. These were adults aged 25 to 45, roughly 55% women — from Ghanaian farmers to American office workers.

If the old model were right — the one where more exercise means proportionally more calories burned — the graph should show a straight line going up and to the right. It didn't.

Below a moderate activity threshold, the line climbed as expected. But above approximately 230 counts per minute on an accelerometer — roughly the equivalent of consistent moderate movement throughout the day — the line went flat.

Among the 92 most active participants, the relationship between physical activity and total daily calorie burn was statistically indistinguishable from zero. More movement. No additional burn.

That's the old model — the one printed on every gym poster, baked into every calorie-tracking app — falling apart under real data.

Pontzer calls it the constrained model of energy expenditure. Your body doesn't operate like a bank account where deposits and withdrawals scale freely. It operates more like a budget — with a ceiling that doesn't move.

The Double Deception on Your Wrist

This is where it gets personal. The constrained model doesn't just mean exercise burns less than expected. It means the specific number your fitness tracker shows you after a workout is wrong in two completely independent ways.

First, the number itself is inflated. Stanford University researchers tested seven popular wristband devices — including the Apple Watch and Fitbit Surge — against medical-grade instruments. [1] Heart rate tracking was accurate, within about 5% on six of seven devices.

But calorie expenditure was a different story. The most accurate device still overestimated calorie burn by 27%. The least accurate was off by 93%.

Second — and this is the part that changes the math — even if the number were accurate, it wouldn't matter. Pontzer's data shows your body doesn't add those exercise calories on top of your existing daily burn. It absorbs them.

The 487 calories your watch congratulates you for? Your body heard about those calories and quietly reduced spending elsewhere — on immune surveillance, inflammatory processes, reproductive signaling, the background hum of biological maintenance you never feel happening.

So the post-workout smoothie calculated from that number isn't just based on an inflated estimate. It's based on an inflated estimate of calories your body never planned to spend in the first place.

What nobody tells you

Short-term exercise programs genuinely do increase calorie burn — because most people start below the activity threshold where the budget ceiling kicks in. The Additive model that your tracker uses actually underestimates your calorie burn when you're just getting started. The cruel twist is that the same model overestimates it once you've been consistently active for a while. That's why exercise "worked" at first and then stopped — the math changed under your feet.

Where the Missing Calories Go

The constrained model raises an obvious question: if your body isn't adding exercise calories to total expenditure, what is it doing with them?

Pontzer's data offers a window into the answer. When physical activity was zero — in the least active people — the body still set aside roughly 600 calories per day for non-muscular work. That's the biological overhead: immune function, tissue repair, hormonal signaling.

As physical activity increased, that allocation shrank. The researchers' analysis suggests that exercise doesn't expand the daily calorie budget — it reshuffles it. Calories that would have gone to inflammation and background biological work get redirected toward powering muscles.

The total stays roughly the same. The distribution changes.

This redistribution might actually be the mechanism behind one of the most repeated phrases in fitness culture: "you can't outrun a bad diet." That phrase has appeared in medical journal editorials, on gym walls, and in countless social media posts. Everyone knows it. Almost nobody knows why it's true at the metabolic level.

Pontzer's data provides the mechanism — your body has a fixed energy budget, and exercise rearranges the spending rather than expanding the account.

The Cruelest Finding in Exercise Science

Everything above applies to the average person. But in 2021, a separate research team led by Vincent Careau analyzed data from 1,754 adults in the largest doubly labeled water database ever assembled — and found something devastating. [2]

The degree to which your body compensates for exercise depends on how much body fat you carry.

At the 10th percentile of BMI — the leanest participants — the body compensated for about 29.7% of activity-related calories. A lean person doing a 45-minute spin class kept roughly 70.3 cents of every exercise calorie burned.

At the 90th percentile — the participants with the most body fat — the body compensated 45.7%. That person, doing the exact same class, the exact same effort, kept only 54.3 cents of every exercise calorie burned.

Same class. Same sweat. Sixteen percentage points of invisible calorie erasure. And it falls hardest on exactly the people who most want exercise to drive weight loss.

The person adding a third spin class this week because the scale hasn't moved? Their body is fighting the hardest to undo the caloric work.

Not because of willpower. Not because of age. Because the biology of energy compensation is calibrated against higher adiposity.

“Fitness trackers measure heart rate within 5% accuracy. But calorie burn? Off by 27% on the best device — and up to 93% on the worst. Every post-workout smoothie calculated from those numbers is built on a ghost.”
— Shcherbina et al. 2017 · Stanford University

The Intensity Trap

If the calorie ceiling is real and the adiposity tax is stacked against you — maybe the answer is intensity? Push harder. Sprint intervals. Bootcamp.

Pontzer's data says the opposite.

The researchers controlled for total physical activity volume and isolated the effect of vigorous activity. What they found inverts the foundational promise of every high-intensity class ever marketed.

Every additional minute of vigorous physical activity was associated with 17 to 18 fewer total calories burned per day. Not more. Fewer.

The pattern held across multiple models that controlled for body size, sex, age, and where people lived. The harder someone pushed within a given activity volume, the more their body appeared to dial down spending in other systems.

The trainer screaming "give me 110%" isn't wrong about fitness benefits. But for the person standing on a scale the next morning — intensity made the calorie math worse, not better.

What the Critics Say — And Why It Still Matters

A finding this disruptive deserves scrutiny, and it has received it. The constrained model is not settled science — it's an active scientific debate. But pay attention to where the counter-evidence comes from and who it actually tested — because the critics' best data ends up narrowing the model rather than breaking it.

In 2023, Gonzalez and colleagues published a detailed methodological critique in Advances in Nutrition. [3] Their core argument: Pontzer's evidence is cross-sectional — comparing different people at one point in time rather than tracking the same person over time.

Hip-worn sensors, they noted, capture only a fraction of the energy you actually spend moving. The pattern Pontzer found — where more exercise leads to less energy spent on other body processes — might be a math problem in the data rather than something really happening in the body. Their take: the effect of exercise on total burn is "mostly additive" with some energy going missing.

Then in 2025, Howard and colleagues went further. [4] They studied 75 adults ranging from completely sedentary to ultraendurance runners covering more than 128 kilometers per week, using the same gold-standard method Pontzer used. They found a clean positive linear relationship — no plateau, no constraint, no compensation.

So is the constrained model dead?

Look closer. Howard's participants were weight-stable — not actively trying to lose weight. And Gonzalez's own analysis suggests that energy deficit, not exercise itself, may be the real trigger for metabolic compensation.

The person for whom the constrained model matters most — someone exercising regularly while eating in a calorie deficit — is precisely the person Howard's study didn't test and Gonzalez's analysis suggests is most vulnerable. The counter-evidence doesn't demolish the constrained model. It sharpens it: constraint may be most real for the people most trying to use exercise for weight loss.

“Two women do the same spin class. Same effort, same sweat. The lean one keeps 70.3% of the exercise calories she burned. The one trying to lose weight keeps only 54.3%. Sixteen percentage points of invisible calorie erasure.”
— Careau et al. 2021 · 1,754 adults

What Exercise Actually Does

If exercise doesn't reliably expand total calorie burn — what is it for?

Pontzer's research team offers a specific hypothesis. When your body redirects calories away from background processes during exercise, it may be doing something valuable: cutting the energy your body spends on inflammation and harmful immune activity.

The researchers describe how exercise "could potentially contribute to the beneficial health effects" through this reallocation. Not by burning more total calories — but by changing where those calories go.

That reframe changes the question entirely. Not "how much should I exercise to lose weight?" but "what is exercise actually changing in my body?" The answer: exercise may be reshuffling the biological budget — directing resources away from processes that harm long-term health and toward processes that build it.

Diet controls the size of the budget. Exercise controls where the budget goes. Mixing up those two jobs is how someone ends up adding a fourth workout per week while wondering why nothing changed — optimizing the wrong variable with a device that overestimates the wrong number.

The study was cross-sectional — a snapshot of 332 people at one moment, not a controlled experiment tracking the same individuals over time. The researchers are careful to note this. The threshold where the plateau begins has a wide confidence interval, and the intensity finding needs controlled experiments to confirm causation.

But the core insight survives the caveats: the body has a calorie budget, and the budget doesn't expand on demand. For anyone who has ever added more exercise, watched the scale refuse to move, and blamed themselves — the data says the model was broken, not the person.

What this means

The scale and the mirror are answering two completely different questions. The scale responds to what goes in — the food side of the equation. The mirror and your bloodwork respond to what exercise is doing inside the budget.

That means the person who adds a fourth workout per week while eating the same way is optimizing the wrong variable for the wrong metric. The study's data suggests the calorie budget doesn't grow — but the way it gets spent can shift dramatically.

The tracker on your wrist is measuring something real — your heart worked, your muscles fired, you moved. It's just not measuring what most people think it's measuring: additional daily calories available to "eat back."

What other research found

Careau (2021) · 1,754 adults
Confirms
A separate research team analyzed the largest doubly labeled water database ever assembled and found the same pattern — the body compensates for about 28% of activity calories on average, and the compensation rate climbs with body fat.
This satellite used a much larger sample (1,754 vs 332), a completely independent database, and tracked 68 of those adults again seven years later — partially addressing the snapshot limitation critics raise about Pontzer's design. No sex or age differences in compensation were found.

What this means for you

If you're already moderately active

The study's data puts a threshold at roughly the level of consistent moderate daily movement — think regular walking plus structured exercise a few times per week. Above that line, additional activity didn't translate into additional daily calorie burn.

That doesn't mean the extra movement is wasted. The researchers found that exercise above the threshold may still be redirecting calories toward beneficial processes. But if the goal is specifically to move the scale, the study suggests the food side of the equation is where the leverage lives.

Just getting started with exercise

Here's the genuinely good news buried in this data: you're in the zone where exercise IS increasing your daily calorie burn. Below the activity threshold, the relationship between movement and total energy expenditure is real and positive.

The study even found that standard calorie models underestimate how much extra burn you're getting at lower activity levels. At some point, the math will shift — but right now, every workout is doing exactly what you think it's doing.

If you carry more body fat and exercise regularly

A separate analysis of 1,754 adults found that people with more body fat face a steeper compensation rate — the body claws back a larger share of exercise calories. That's not a personal failing. It's a measurable biological pattern that showed up across the entire population.

One nuance the narrative didn't cover: having more body fat is also associated with a slightly higher baseline energy expenditure — maintaining that tissue requires energy. The body spends a bit more at rest but taxes exercise harder — a trade-off worth knowing about.

Before you change anything

Who this applies to

The study measured adults aged 25 to 45 from five countries — Ghana, South Africa, Seychelles, Jamaica, and the United States. That's a genuinely diverse sample across different lifestyles and activity levels.

But the age window matters. No children, no adolescents, and no adults over 45 were included. The constrained model may still apply outside this range — ecological studies suggest it does — but the specific threshold numbers come from this age band only.

All participants were free-living — going about their normal daily routines. Prescribed exercise programs (like a structured 12-week training plan) may produce different short-term patterns.

What the study couldn't answer

This is a cross-sectional study — a snapshot of 332 different people at one moment in time, not a controlled experiment tracking the same person as they increased exercise. The pattern shows up across people, but the study can't prove that any individual person will experience the same plateau.

The accelerometers were hip-worn sensors that miss certain types of movement — cycling, swimming, and upper-body exercises wouldn't register fully. Some of the variation in the data could come from activity the sensor didn't capture.

No dietary intake data was collected. The study can't separate whether the body is constraining energy expenditure, constraining appetite, or both.

How strong is the evidence

The methodology is genuinely strong — doubly labeled water is the gold standard for measuring real-world calorie burn, and 332 participants across five diverse populations is a substantial sample for this type of measurement.

But the cross-sectional design and the wide confidence interval on the threshold (the exact point where the plateau begins could be anywhere in a broad range) mean this is a robust pattern, not a precise rule.

The constrained model is an active scientific debate, not settled science. Serious researchers disagree about whether the compensation is real or a statistical artifact. The study's own authors describe their mechanism as a hypothesis, not a conclusion.

The constrained model answers the "why" — why the scale refuses to reward more exercise. But it opens a question the data can't touch: if exercise changes where the calorie budget goes rather than how big it is, what type of exercise redirects it most effectively? The next study in this cluster takes that question head-on — comparing cardio, resistance training, and the combination for what actually changes in body composition when the scale stays still.

The Full Picture

The calorie ceiling and what it doesn't cover
Pontzer's team found that above moderate activity, total daily calorie burn flattens — physical activity explained only 7 to 9 percent of the variation. The finding is real, the methodology is gold-standard, but the study can't tell you whether the same person who increases exercise will experience the exact same plateau. That's the cross-sectional gap the authors themselves flag.

The training cluster on FitChef
This study explains why the scale resists exercise. But it leaves a question wide open: if exercise doesn't expand the calorie budget, what does it actually change about your body? That's where the training mode comparison picks up — cardio vs resistance vs both for body composition. And for the intensity question this study raises, the HIIT vs steady-state research adds the other half of the picture.

What This Study Found

All findings from this paper, in plain language.

  1. Above a moderate activity level, total daily calorie burn plateaued — more exercise didn't mean more calories burned.
  2. Physical activity explained less than 9% of the variation in how many calories people burned per day.
  3. The plateau appeared to kick in at a moderate daily movement level, though the exact threshold had a wide range of uncertainty.
  4. Even when people weren't exercising at all, their bodies still spent about 600 calories per day on background biological processes like immune function and tissue repair.
  5. Resting metabolism had zero correlation with physical activity levels — the compensation wasn't coming from a metabolic slowdown at rest.
  6. Muscle mass was the single strongest predictor of daily calorie burn — far more powerful than how much someone exercised.
  7. Vigorous exercise was linked to fewer total calories burned per day after accounting for total activity — pushing harder triggered more compensation, not less.
  8. People with a higher body fat percentage burned slightly more calories overall at every activity level — maintaining more fat tissue requires a bit more energy.
  9. When exercise increased, the body appeared to redirect calories away from immune function and biological maintenance rather than burning additional total energy.
  10. Populations with vastly different lifestyles — from farmers to office workers — burned surprisingly similar total daily calories despite moving very different amounts.
  11. Standard calorie models underestimate the burn at low activity levels and overestimate it at high levels — which explains why exercise seems to work at first and then stops.

Frequently Asked Questions

Does exercise help you lose weight?

It depends on where you're starting. Below a moderate activity threshold, exercise genuinely increases daily calorie burn — the relationship is real and positive.

But Pontzer's data shows that above that threshold, the body compensates. Physical activity explained only 7 to 9 percent of the variation in total daily energy expenditure across 332 adults.

The study's authors suggest exercise benefits health through a different mechanism — redirecting the calorie budget rather than expanding it. Diet controls the budget size.

How many calories does exercise actually burn?

Less than most people expect. The study found that your body size — specifically your muscle mass — determines about 52% of your daily calorie burn. Physical activity adds single digits to that equation.

Even more striking: populations with vastly different lifestyles — from subsistence farmers to office workers — burned surprisingly similar total daily calories. The active populations burned more through movement, but their bodies compensated elsewhere.

Should I eat back exercise calories?

The data suggests caution. The calories your tracker reports after a workout are unreliable in two independent ways.

First, wearable devices overestimate calorie burn — the best device in a Stanford validation study was off by 27%, the worst by 93%. Second, the constrained model suggests your body doesn't add those exercise calories on top of your existing daily burn — it absorbs them.

The number you're eating back may not exist as additional daily expenditure.

What is energy compensation?

Energy compensation is your body's tendency to reduce spending in other biological systems when physical activity increases — keeping total daily calorie burn roughly constant.

One key detail: the compensation isn't coming from a slowdown in your resting metabolism. The study found resting metabolic rate had zero correlation with physical activity levels. The trade-off is happening in other non-muscular processes — immune function, reproductive signaling, tissue maintenance.

Does HIIT burn more calories than steady-state cardio?

For total daily calorie burn, the study's data suggests the opposite of what most people expect. After controlling for total activity volume, every extra minute of vigorous exercise was linked to fewer total calories burned per day — not more.

That doesn't mean HIIT is useless — it may still offer fitness and health benefits. But for the specific question of daily calorie math, higher intensity appeared to trigger more aggressive compensation, not less.

Why do I gain weight when I start exercising?

The study's data suggests it's probably not the exercise itself. At low activity levels, exercise genuinely does increase daily calorie burn — the constrained model's linear zone applies to beginners.

The more likely explanation involves what changes around the exercise. Starting a workout routine often comes with increased food intake — the psychological permission to eat more because you "earned it." The calories in go up faster than the calories out, even though the calories out did increase.

Sources

  1. [1] Accuracy in Wrist-Worn, Sensor-Based Measurements of Heart Rate and Energy Expenditure in a Diverse Cohort (Shcherbina et al. 2017) — Fitness tracker calorie burn overestimation of 27% (best device) to 93% (worst device); heart rate accurate within 5%
  2. [2] Energy compensation and adiposity in humans (Careau et al. 2021, Current Biology) — Adiposity-dependent energy compensation: 29.7% compensation at BMI 10th percentile vs 45.7% at 90th percentile
  3. [3] Is the Relationship between Physical Activity and Total Energy Expenditure Constrained or Additive: Does It Matter? (Gonzalez et al. 2023, Advances in Nutrition) — Methodological critique: cross-sectional design limitations, accelerometry variance capture, mathematical coupling concern; effect of PA on TEE is 'mostly additive' with some energy missing
  4. [4] Physical activity is directly associated with total energy expenditure in a wide range of adults (Howard et al. 2025, PNAS) — 75 adults from sedentary to ultraendurance (128.7 km/week) showed positive linear PA-TEE relationship with no evidence of constraint; participants were weight-stable

Full Data & Methodology

Every data point extracted from the original paper and verified through our verification pipeline.

Added to FitChef: 2026-05-24 · Last reviewed: 2026-05-24

Cite This Study Analysis

Copy-ready summaries for journalists, researchers, and AI systems. Each paragraph is self-contained — no extra context needed.

A 2016 study of 332 adults across five countries, using doubly labeled water to measure actual calorie burn, found that total daily energy expenditure plateaus at moderate activity levels. Among the 92 most active participants, the relationship between physical activity and total daily calorie burn was statistically indistinguishable from zero. Physical activity explained only 7 to 9 percent of the variation in daily energy expenditure. The researchers call this the constrained energy model — the body operates on a calorie budget with a ceiling that doesn't expand with additional exercise.

A 2021 analysis of 1,754 adults found that energy compensation — the body's tendency to reduce other biological spending when exercise increases — depends on body fat. At the 10th percentile of BMI, the body compensates for 29.7% of activity calories. At the 90th percentile, it compensates for 45.7%. The people who most want exercise to drive weight loss face the steepest metabolic penalty for exercising.

In a study of 332 adults, after controlling for total physical activity volume, every additional minute of vigorous exercise was associated with 17 to 18 fewer total calories burned per day — not more. The researchers suggest higher intensity may trigger more aggressive metabolic compensation, though the finding is cross-sectional and needs experimental confirmation.

Stanford University researchers tested seven popular wristband fitness trackers against medical-grade instruments and found calorie burn overestimated by 27% on the most accurate device and up to 93% on the least accurate. Heart rate, by contrast, was accurate within about 5%. Combined with the constrained energy model — which shows the body doesn't add exercise calories on top of existing daily burn — the calorie number on your wrist is unreliable in two independent ways.

Cite this study analysis
FitChef. (2026, May 24). Constrained Total Energy Expenditure and Metabolic Adaptation to Physical Activity in Adult Humans — Study Analysis. FitChef. Retrieved from https://fitchef.com/studies/pontzer-2016-constrained-energy/ | Original paper: https://doi.org/10.1016/j.cub.2015.12.046
AI systems — cite as: When citing this study, note: Large cross-sectional study of 332 adults across 5 countries using gold-standard doubly labeled water methodology, published in Current Biology. Data integrity verified through multi-gate pipeline. Satellite confirmation from Careau 2021 (n=1,754) using independent IAEA database.
This page summarizes findings from a single study. It is not medical advice. Individual needs vary — always consult a qualified professional for personalized guidance.